Hierarchical model predictive control of greenhouse energy systems considering energy-water-carbon-food nexus

Greenhouse cultivation plays a vital role in ensuring food security but is often associated with high energy consumption, water usage, and carbon emissions. Integrating renewable energy systems for power supply and utilizing rainwater harvesting for irrigation can help address these challenges. However, balancing these interconnected factors requires advanced control strategies. In this study, we propose a hierarchical model predictive control (MPC) framework to optimize the management of grid-connected photovoltaic-battery systems in greenhouses, accounting for the interactions among energy use, water consumption, carbon emissions, and food production (EWCF nexus). The hierarchical MPC is structured in three layers: the first optimizes greenhouse operations to minimize total costs (MTC); the second manages the scheduling of the hybrid energy system to minimize operational cost (MOC); and the third designs an MPC controller to handle photovoltaic generation and load demand disturbances. Results show that the proposed MTC strategy reduces the total cost by 81.01% compared with the minimizing energy consumption strategy. Moreover, the MOC strategy reduces operational costs by 20.68% compared to the maximizing self-consumption strategy. In addition, the proposed MPC achieves superior performance in tracking the reference trajectory under varying disturbance levels compared to commonly used open loop controllers. This study provides practical guidance for greenhouse management by addressing key resource and environmental challenges, contributing to the sustainable development of controlled-environment agriculture.